Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for inter-table parallel refresh in a database management system (DBMS), the method comprising: receiving, by one or more computer processors, DBMS source database and DBMS target database wherein the DBMS source database comprise source parent datasets and related source child datasets; determining, by one or more computer processors, a collection of the source parent datasets and the related source child datasets for the inter-table refresh, wherein determining the related source child datasets is based on matching a source child dataset key field identifier with at least one of a source parent dataset key identifier or a source parent dataset foreign key identifiers; creating, by one or more computer processors, a collection of parent refresh segments and a related collection of child refresh segments based on the collection of the source parent datasets and the related source child datasets, wherein each parent refresh segment of the collection of parent refresh segments comprises a group of rows subdividing the source parent datasets, and wherein each child refresh segment of the related collection of child refresh segments comprises a related group of rows subdividing the source child datasets; creating, by one or more computer processors, a collection of parent refresh batches and a collection of child refresh batches based on the collection of parent refresh segments and the collection of child refresh segments respectively; outputting, by one or more computer processors, an inter-table refresh batch schedule based on the collection of the parent refresh batches and the collection of the child refresh batches, wherein the collection of the parent refresh batches and the collection of the child refresh batches operate concurrently and a current child refresh batch follows a completion of a respective current parent refresh batch; identifying, by one or more computer processors, a sub-child wherein the sub-child is based on a first source child dataset being related to a second source child dataset wherein a first source child dataset foreign key identifier matches a second source child dataset key identifier; responsive to identifying the sub-child, marking, by one or more computer processors, the first source child dataset as a source parent dataset and marking the sub-child as a source child dataset for recursive processing of the inter-table refresh until the found sub-child is segmented; and executing, by one or more computer processors, the inter-table refresh batch schedule.
This invention relates to database management systems (DBMS) and addresses the challenge of efficiently refreshing related datasets across multiple tables in parallel. The method involves a source database containing parent and child datasets, where child datasets are linked to parent datasets via key or foreign key identifiers. The system first identifies all related parent and child datasets for the refresh operation by matching key fields. It then subdivides these datasets into segments—parent refresh segments for parent datasets and child refresh segments for child datasets—where each segment contains a group of rows. These segments are further grouped into parent and child refresh batches. The system generates a refresh schedule where parent and child batches operate concurrently, with child batches executing only after their corresponding parent batches complete. The method also handles nested relationships by recursively identifying sub-child datasets (where a child dataset references another child dataset) and reprocessing them as parent-child pairs until all relationships are segmented. Finally, the system executes the scheduled refresh batches, ensuring efficient and parallelized updates across interdependent tables. This approach optimizes performance by minimizing dependencies and maximizing parallel processing during database refresh operations.
Unknown
June 9, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.